Abstract
Metagenomics is the study of environmental microbial communities and has various applications and implications in biological research. This paper aims to study the role of microbial communities in cattle rumen and their relation to probiotic diet supplement usage as part of the EU H2020 MetaPIat project l1 MetaPlat, http://www.metaplat.eu. In this research, we proposed and evaluated a computational framework to classify 16S rRNA samples from Bos taurus (cattle) rumen microbiome into a diet phenotype. We performed analysis by benchmarking various phylogeny-driven methods based on integration of biological domain knowledge of relationships and non-phylogenetic methods based on the raw abundances. The integrative approach incorporating phylogenetic tree structure into machine learning (ML) modelling achieved a high predictive performance with Accuracy of 0.925 and Kappa of 0.900 for classifying cattle microbiomes into diets supplemented with oil, nitrate, a combination and controls.
Original language | English |
---|---|
Title of host publication | Proceedings - 2018 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2018 |
Editors | Harald Schmidt, David Griol, Haiying Wang, Jan Baumbach, Huiru Zheng, Zoraida Callejas, Xiaohua Hu, Julie Dickerson, Le Zhang |
Place of Publication | Madrid, Spain |
Publisher | IEEE |
Pages | 1854-1860 |
Number of pages | 7 |
ISBN (Electronic) | 9781538654880 |
ISBN (Print) | 978-1-5386-5489-7 |
DOIs | |
Publication status | Published (in print/issue) - 24 Jan 2019 |
Keywords
- Phylogeny
- Cows
- Radio frequency
- Predictive models
- Vegetation
- Pipelines
- Tools
- machine Learning
- classification
- metagenomics
- ordination
- Cattle Microbiomes
- phylogeny